3. Pragmatics and language teaching
3.3 The teachability of pragmatics
3.3.1 Explicit and implicit teaching approaches
Discussions over the benefits of explicit versus implicit teaching approaches occupy
much of the debate concerning the effectiveness of instructed SLA. One of the
underlying issues to be addressed in the design of interventional studies is the choice
of the explicit or implicit dichotomy, which is differentiated by the presence (explicit)
or absence (implicit) of metapragmatic information as part of the instructional input
learners have adopted an explicit teaching approach (Jeon & Kaya, 2006, Takahashi,
2010; Taguchi, 2015), which is often characterised by teacher-led introduction of the
pragmalinguistic and sociopragmatics goals of the target language. Activities to
promote learning in explicit treatments include awareness-raising tasks and activities
providing communicative practice such as role plays (Kasper, 1996; Safont Jorda,
2004).
Overall, findings do show students having profited from explicit instruction (e.g.,
Bouton, 1994; Cohen & Tarone, 1994; Fukuya, 1998; Iwai, 2013; Kondo, 2010; Liddicoat & Crozet, 2001; Lyster, 1994; Uso-Juan, 2013; Wishnoff, 2000). Studies
specifically targeting request and apology language also follow this positive trend;
Eslami and Eslami-Rasekh, (2008, requests and apologies); Halenko and Jones,
(2011, requests); Johnson and deHaan, (2013, requests and apologies); Martinez-
Flor, (2008, requests), Safont Jorda, (2004, requests). Safont Jorda (2004) adopted a
pre-posttest measure with 160 beginner-immediate level Spanish undergraduate
students on an EAP course. Following one semester of explicit instruction targeting
linguistic forms in oral and written requests, gains from two oral and written
production tasks were analysed according to the amount and type of request head
acts employed. Findings revealed an increase in quantity and type of request head
act produced, post-instruction. Specifically, a higher frequency of conventionally-
indirect strategies and fewer direct strategies were reported. Low level Spanish undergraduate EFL students were also the sample employed in Martinez-Flor’s
(2008) study though the pre-posttest design here examined the frequency and type of
internal and external request modifiers. Six hours of explicit treatment were
operationalised for 38 students through phased sessions consisting of awareness-
raising and production activities. Positive instructional effects were again reported,
instances of fixed expressions. Neither study incorporated delayed tests to measure
long term recall or attrition rates.
Halenko and Jones (2011)3 adopted a similar methodological approach with six hours
of instruction on requests, emphasising pragmalinguistic and sociopragmatic
awareness, consolidated by substantial communicative practice. What differentiates
the Halenko and Jones study is the British ESL setting, the inclusion of a control
group and delayed test to determine short and long term instructional effects, and
focus group data on the benefits of pragmatic classroom input. Positive post- instructional effects were clearly evident. According to NS raters, there were
significant pretest-posttest improvements on request responses to a six-item
production task, though these declined following a six-week period. The authors
propose that even a short term focus on pragmatics embedded within existing
language programmes can be beneficial, as also reflected in the learner interview
comments. However, the authors suggest that sustained practice is required for long
term retention of pragmatic request knowledge.
According to the most recent review of intervention studies employing designs
suitable for measuring true explicit instructional effects (Taguchi, 2015), only two
studies include a focus on apology language. First, Eslami and Eslami-Rasekh (2008)
examined the teaching of both requests and apologies to an underexplored group of
Iranian postgraduate non-native English speaking teacher candidates (NNESTCs)
(experimental group = 25, control group = 27). Following seven hours of instruction,
the responses from an error recognition task (judging pragmatic awareness) and a
written production task (judging pragmatic production) were assessed for appropriacy
on a five-point Likert scale. Despite any specific details as to the reason for
improvements, the treatment was declared to be highly effective for both production
3 Some of the instructional material designed by the author of this present study was replicated in the training materials of Halenko and Jones (2011) to assess its effectiveness.
and awareness of request and apology language, based on the raters’ pretest and
posttest scores. For Eslami and Eslami-Rasekh, the instrumental motivation as pre-
service English teachers to engage in the intervention was seen as the primary
influence on the results.
Secondly, Johnson and deHaan’s (2013) study also developing request and apology
language with 22 undergraduate computer science students, illustrates the potential
of technology-enhanced instruction (see section 3.5 for further discussion of
technology-based pragmatic input). Utilising model conversations, a classroom online wiki space and digital video software, learners were able to perform, record, self-
correct and reflect on their pragmatic outputs within business contexts, under the instructor’s guidance. Pretest-posttest NS assessments revealed improved levels of
accuracy and appropriacy for request and apology language, post-instruction.
Statistical differences were only noted in appropriacy, however, realised through
fewer direct, simplistic formulations (as found in the pretest), replaced by increased modality, conventionally-indirect language and fixed expressions such as, ‘I was wondering if’, after the treatment. For both the Eslami and Eslami-Rasekh, and Johnson and deHaan studies, delayed-posttests were not administered, nor were the
value or effectiveness of teaching multiple speech acts investigated.
Explicit instruction has not consistently been found to be superior, as several studies
comparing the explicit and implicit approaches have found. Kubota (1995), for
instance, found initial gains from explicit instruction had disappeared by the time a delayed posttest was employed, Rose and Ng’s (2001) pre- and posttests did not
produce positive results on all of the assessment measures employed, and Martinez-
Flor (2006) reported similar levels of effectiveness for both explicit and implicit
treatments. Variability in operationalising these two methods is a suggested cause of
positive results in their own right so this instructional approach cannot be entirely
dismissed (e.g., Fukuya & Zhang, 2002; Takimoto, 2009).
The implicit condition is generally characterised by the learners’ inductive self-
discovery of the target language features, designed to raise awareness. Fukuya and
Zhang (2002) investigated the effects of implicit corrective feedback in the form of
pragmatic recasts with a treatment and control group. The 20 intermediate level L1
Chinese speakers participated in seven 50-minute sessions of role play practice of
request language, enhanced by the inclusion of pragmatic recasts for the treatment group. Results from a written production task of the same role play items revealed the
treatment group used significantly more target-like forms than the control group. The
authors attributed the noticeable effects of the implicit treatment to the recasts encouraging learners’ to ‘notice’ the gaps between the interlanguage and target
language forms. For Fukuya and Zhang, this was also aided by the formulaic nature of the recasts e.g., ‘would you mind’, ‘I was wondering if’, which they claim are expressions which can be easily stored as patterns. Takimoto (2009) furthermore attributed the ‘noticing’ hypothesis to findings that treatments involving the presence
or absence of explicit instruction were equally effective. Following 40 minutes of
treatment on polite request forms, no significant posttest or delayed-test differences
were noted between the treatment groups, but all significantly outperformed a control
group on a series of oral, aural and written tests.
What links the success of the implicit treatments in the above studies is that simple exposure is not sufficient. There is a need to first ensure learners’ attention, and then
direct this to noticing and to subsequent processing of the information to induce rules,
given no overt explanation of the target features is available in this approach. This indicates, regardless of teaching method, Schmidt’s (1993) hypothesis of noticing and
processing information still needs to be observed. Put simply, when implicit instruction versus no-instruction designs are investigated, treatment gains are
generally exhibited, but performances by explicitly-instructed groups generally
surpass those receiving implicit treatment (Taguchi, 2010 provides a comprehensive
review). Given the general consensus of the superiority of explicit over implicit
treatments, the explicit teaching approach was adopted for the instructional phase of
this thesis.
The reported variability of instructional success may also be attributable to an
interplay between instructional method and other external factors. Firstly, research
evidences pragmatic instruction should be at least five hours to be effective (Jeon & Kaya, 2006; Salazar, 2003; Uso-Juan, 2013). This is in contrast to Norris and Ortega’s (2000) synthesis of grammar-focussed instructional studies which found
shorter treatments of up to three hours were more beneficial. An important
differentiation within target features is raised here, suggestive that perhaps L2
pragmatic instruction requires more attention through longer instructional periods
than other language features such as grammar due to: i) the subconscious nature
with which pragmatic knowledge is acquired by NS, ii) the fact that it has no codified
rules, iii) the traditional emphasis placed on grammatical knowledge over pragmatic
knowledge in the classroom, and iv) the fact that it is given little attention in language
coursebooks. Secondly, individual learner differences such as proficiency (e.g.,
Codina-Espurz, 2008) and motivation (e.g., Takahashi, 2012) are also reported to
have an impact on pragmatic development. The claim that motivation, for instance,
may be restored or enhanced through use of technology as a medium for learning, is
a key aspect of the present study. A recent innovation in pragmatics instruction has
seen the incorporation of technology-enhanced learning and assessment materials
for language development: a trend explored further in section 3.5. Finally, in addition to teachability, ‘learnability’ of the target feature is said to affect instructional success
(Taguchi, 2010). For instance, Johnson and deHaan (2013) reported greater
apology production than at the micro-level when measuring accuracy of response. In
other words, politeness strategies and discourse moves were more easily retained
and recalled from the instruction than knowledge of the linguistic form. A similar trend was evidenced in Sykes’ (2009, 2013) online studies into developing request and
apology behaviour. Findings revealed minimal change in choice of request strategies,
in contrast to clear improvements in some aspects of apology language such as head acts. Sykes claims the structural and functional simplicity of apology formula at the
lexical level facilitated learning. Findings such as these widen the debate concerning
the benefits of the teaching and learning of formulaic language: an area discussed
further in section 3.4.
In summary, several empirically-tested instructional frameworks and numerous offline
and online practice activities now exist to help develop successful teaching
programmes. In addition, the overwhelming evidence that explicit teaching
approaches produce the most effective results provides further guidance for
maximising learning outcomes. Still, the number of instructional studies falls short of
comparative investigations and those adopting an instruction versus exposure
approach are still heavily underexplored. The present study aims to fill that gap. The
next sections assess the value of adopting formulaic language (3.4) and technology-
based training materials (3.5) for classroom interventions, which are key features of
the present study.